This is a revised renewal application for our ongoing case-control study of nutritional and other risk factors for adenomatous polyps of the large bowel. Subjects are asymptomatic persons who attended a sigmoidoscopy screening clinic at one of two Kaiser Permanente-Southern California clinical centers. Cases (n=530) are persons with a first-time diagnosis of at least one adenomatous polyp confirmed by biopsy. Controls (n=530) are subjects found to have no adenomatous polyp at sigmoidoscopy and with no history of large bowel neoplasia. Data are obtained from a food frequency questionnaire, in-person interview, fasting blood sample, and pathology material (including the recent addition of tumor blocks). In the current study, we have exceeded the sample size goal and achieved very high response rates (84% for cases, 82% for controls) for a study of this nature. Our objective for this renewal is to investigate gene-environment interactions, including both "inherited susceptibility genes" and "metabolic genes". During the next grant period, we plan to focus on a set of mutator genes (hMSH2, hMLH1, PMS1, PMS2), NAT1, NAT2, and GST. hMSH2 is a relatively common gene believed to be a major cause of HNPCC and other forms of colon cancer/polyps. Molecular analyses of these genes (principally looking for microsatellite instability and sequencing) are planned that would enable us to a) estimate the prevalence of microsatellite instability in polyps b) the number and type of mutations in hMSH2, hMLH1, and PMS, c) the prevalence of these mutations. d) the penetrance/risk associated with mutations, and e) identify factors that may affect penetrance of these genes (including other genes, such as k- ras, and environmental exposures). NAT helps determine acetylator status. We propose to determine genotypes of NAT1 and NAT2. We will also genotype GST1. This will enable us to estimate main effects for NAT and GST, and interactions with protein intake in general, red meat, and tobacco smoke. In order to achieve adequate statistical power to detect interactions. we propose to double our sample size to 1200 cases and 1200 controls, using essentially the same staff and protocol as our ongoing study. Achieving this sample size would also enable us to begin more informative explorations of nutrient-nutrient and physical activity-nutrient interactions.